轉載來源https://blog.csdn.net/weixin_43569478/article/details/83745105
GSEA和常規的GO、KEGG的差異在于，GSEA使用的是基因集?；蚣牟町惙治?。GSEA可以清除的說明上調基因和下調基因分別參加了哪些通路，生物學途徑。
GO、KEGG只能給出差異基因的pathway,生物學通路、表達部位等。無法卻分哪部分是上調基因，哪部分是下調基因。

同時說明Y叔的clusterProfiler的包中的gseGO和gseKEGG的genelist個格式。
https://github.com/GuangchuangYu/DOSE/wiki/how-to-prepare-your-own-geneList

gsea的原始數據分為三列，一列是geneid,一列是FoldChange,一列是根據FoldChange排序的結果。

其實gsea就是看foldchange的值的分布，如果是隨機分布，那結果就不理想。我們想要的是在兩端富集分布。
GSEA分析同樣可以使用clusterprofiler包。Y叔的包真的強大，我還要用它做轉錄因子的富集分析。

GSEA的數據集是所有基因的entrez id，和Log2FoldChange的值。不是設置p-value<0.05和log2FoldChange>1過濾后的值。

library(DO.db)
require(DOSE)
library(clusterProfiler)
library(AnnotationHub)
library(readr)
library("genefilter")
library(pheatmap)
library(tidyverse)
library(DESeq2)
library(ggplot2)
library(export)
library(enrichplot)
library(Rgraphviz)
#構造圖片輸出函數，need input filename width height
#函數依賴export包
out_img <- function(filename,pic_width=5,pic_height=7){
  graph2png(file=filename,width=pic_width,height=pic_height)
  graph2ppt(file=filename,width=pic_width,height=pic_height)
  graph2tif(file=filename,width=pic_width,height=pic_height)
}
#all_entrez.csv的第一列是entrezid,第二列是FoldChange的值。
#gse需要單獨做數據格式
d <- read.csv("all_entrez.csv")
geneList <- d[,2]
names(geneList)=as.factor(d[,1])
geneList <- sort(geneList,decreasing = TRUE)

#gseGO進行GSEA分析
#參考連接https://yulab-smu.github.io/clusterProfiler-book/chapter12.html
###gseBP <- gseGO(geneList=geneList,ont="BP",OrgDb=maize,keyType = 'ENTREZID',nPerm = 50000,minGSSize = 100,maxGSSize = 6000,pvalueCutoff = 0.05,verbose = FALSE)

############# GSEA CC 模式 start
ego3 <- gseGO(geneList = geneList,OrgDb = maize,ont = "CC",nPerm = 1000,minGSSize = 100,maxGSSize = 1000,pvalueCutoff = 0.05,verbose = FALSE)
write.csv(ego3,file = "GESA-GO_CC.csv")
#ridgeline plot for expression distribution of GSEA result
ridgeplot(ego3)
out_img(filename = "ridgeplot_CC",pic_width = 12,pic_height = 12)
#只顯示值最高的一組的信息
#gseaplot(ego3,geneSetID = 1,by="runningScore",title=ego3$Description[1])
#gseaplot(ego3,geneSetID = 1,by="preranked",title=ego3$Description[1])
#gseaplot(ego3,geneSetID = 1,title=ego3$Description[1])

#顯示前4組信息
gseaplot2(ego3,geneSetID = 1:4, ES_geom = "dot",pvalue_table = TRUE)
out_img(filename = "gseaplot_CC",pic_width=12,pic_height = 10)

#gsearank(ego3,1,title=ego3[1,"Description"])
############GSEA CC 模式end

############# GSEA BP 模式 start
ego2 <- gseGO(geneList = geneList,OrgDb = maize,ont = "BP",pvalueCutoff = 0.05,verbose = FALSE)
write.csv(ego2,file = "GESA-GO_BP.csv")
#ridgeline plot for expression distribution of GSEA result
ridgeplot(ego2)
out_img(filename = "ridgeplot_BP",pic_width = 12,pic_height = 12)
#只顯示值最高的一組的信息
#gseaplot(ego3,geneSetID = 1,by="runningScore",title=ego3$Description[1])
#gseaplot(ego3,geneSetID = 1,by="preranked",title=ego3$Description[1])
#gseaplot(ego3,geneSetID = 1,title=ego3$Description[1])

#顯示前4組信息
gseaplot2(ego2,geneSetID = 1:4, ES_geom = "dot",pvalue_table = TRUE)
out_img(filename = "gseaplot_BP",pic_width=12,pic_height = 10)

#gsearank(ego3,1,title=ego3[1,"Description"])
############GSEA BP 模式end

############# GSEA MF 模式 start
ego4 <- gseGO(geneList = geneList,OrgDb = maize,ont = "MF",pvalueCutoff = 0.05,verbose = FALSE)
write.csv(ego4,file = "GESA-GO_MF.csv")
#ridgeline plot for expression distribution of GSEA result
ridgeplot(ego4)
out_img(filename = "ridgeplot_MF",pic_width = 12,pic_height = 12)
#只顯示值最高的一組的信息
#gseaplot(ego3,geneSetID = 1,by="runningScore",title=ego3$Description[1])
#gseaplot(ego3,geneSetID = 1,by="preranked",title=ego3$Description[1])
#gseaplot(ego3,geneSetID = 1,title=ego3$Description[1])

#顯示前4組信息
gseaplot2(ego4,geneSetID = 1:4, ES_geom = "dot",pvalue_table = TRUE)
out_img(filename = "gseaplot_MF",pic_width=12,pic_height = 10)

#gsearank(ego3,1,title=ego3[1,"Description"])
############GSEA MF 模式end

#gsaKEGG基因富集分析
kk2 <- gseKEGG(geneList = geneList,organism = 'zma',pvalueCutoff = 0.05,verbose = FALSE)
write.csv(kk2,file="GSEA_KEGG.csv")

gseaplot2(kk2,geneSetID = 1:4,ES_geom = "dot",pvalue_table = TRUE)
out_img(filename="GSEA_KEGG",pic_width = 12,pic_height = 10)
ridgeplot(kk2)
out_img(filename="ridgeplot_GSEA_KEGG",pic_width = 12,pic_height = 12)